Acoustic Emission Monitoring

Abstract

Offshore wind turbines are used more and more for the production of our electricity. The wind turbines are located in a remote and harsh environment, and are subject to heavy cyclic loading, which may cause fatigue in the support structures. Periodic inspections are required to assess the structural health of the wind turbines. Acoustic Emission Monitoring is the technique of acquiring and processing of the high-frequency sound waves emitted during crack growth. Accurate processing of these signals can lead to detection and localization of fatigue cracks, and further reduction of the need for costly periodic inspections. The factors that determine the coverage area of acoustic emission sensor nodes are source signal strength, attenuation of signal in the medium, and onsite noise level. Together with a cost benefits analysis, this leads to insight into the feasibility of this technique for offshore wind turbines. Investigation of the attenuation is done using a higher-order Spectral Element Method with the input acquired from experiments. The signal that is detected at the sensor has to have sufficiently higher intensity than the surrounding noise. The most severe sources of noise are rain drops hitting the water surrounding the turbine, for which a setup is used in the laboratory, and noise from rotating equipment, for which a measurement has been performed at an on shore wind turbine. The localization accuracy of 5 – 10 percent was shown to be achievable in a laboratory setup. With the coverage area determined, a monitoring strategy for a single wind turbine is proposed, as well as how an acoustic emission monitoring system can be efficiently implemented at a larger scale for a wind farm. It is concluded that acoustic emission monitoring is generally feasible for this application, yet further testing is required in order to decrease the uncertainties and to demonstrate the capabilities to potential operators.